Simulation of pedestrian flow based on cluster temperature

This paper defines individual temperature of pedestrian flow. After that, it calculates every pedestrian´s individual temperature in each time-step, and uses the mean of individual temperature to express cluster temperature. Simulation of pedestrian flow is based on cellular automata model. In early and middle period, the cluster temperature has significant decrease with evacuation time in the process of evacuation. However, the cluster temperature has obvious characteristics of tailing in later stage. It is shown that the cluster temperature is not only influenced by system scale, but also directly restricted by pedestrian density.